Today (?) in Central Asian history: the Battle of Talas (751)

The Abbasids defeat a Tang Chinese army and secure Islam's foothold in Central Asia.

If you’re interested in history and foreign affairs, Foreign Exchanges is the newsletter for you! Sign up for free today for regular updates on international news and US foreign policy, delivered straight to your email inbox, or subscribe and unlock the full FX experience:

After the early Arab conquests in the seventh century drove the Byzantine Empire out of the Levant and Egypt and drove the massive Sasanian Empire out of existence altogether, the Arabs under the Rashidun and then Umayyad caliphs continued to expand their frontiers. By the early eighth century they’d added vast new territories to their empire. To the west, as we know, the Arabs eventually reached as far as modern Spain before finally stopping at the Pyrenees Mountains. To the southeast, they conquered the region of Sindh (in modern Pakistan) before losing steam as they attempted to continue further east. By the time the Abbasid Revolution ousted the Umayyad dynasty in 750, their empire looked something like this:

To the northeast, in Central Asia, you may notice a marker for “Talas 751” on that map up there, and that’s the battle we’re here to discuss today. It represents the last of three contacts between the expanding Arabs and another expansionist Asian power, Tang Dynasty China. Although Chinese history is not my area, we should say a little something about the Tang here. People for whom Chinese history is their area often describe the Tang period (618-907, with a short break from 690 to 705 when Empress Wu Zetian proclaimed the very short lived second Zhou Dynasty) as a high point—the first half, when combined with the later Song dynasty (10th-13th centuries) of a Chinese “golden age” if you’re into that sort of terminology. The Tang succeeded and built upon the short-lived (581-618) Sui Dynasty, which had reunified China after the division of the “Northern and Southern dynasties” period from the end of the fourth century through the end of the sixth century.

As you might expect, then, the Tang left a considerable legacy for successive Chinese dynasties. The thing that looms largest in this respect is probably the famous imperial civil service examination, which was intended to ensure that merit, not social status, determined who got to serve in desirable bureaucratic positions. As any critic of our present day “meritocracy” could tell you, that’s not really how these things work, but the exam does seem to have been a good faith attempt to open government jobs up to anyone regardless of their parentage or connections at court. Some form of examination had already been in place when the Tang came to power, but they’re credited with making it the criteria for hiring, a status it held until the late Qing dynasty many centuries later. OK, technically I think this shift happened during Wu Zeitan’s interregnum, but the Tang seem to get some of the credit for it anyway.

The Tang are also known for their imperial expansion. By the early eighth century, they directly ruled an empire that extended from the northern part of the Korean peninsula in the east almost to the region of Kashmir in the west, the biggest empire any Chinese dynasty had ruled to that point. If you add territories the Tang controlled indirectly, their dominion was much larger than that. Emperor Taizong (d. 649), the second Tang ruler, used the title “Khan of Heaven” to convey his status not just as the ruler of China, but as the sovereign of numerous other Asian kingdoms that paid homage to him. Tang emperors received tribute from places like Tibet, Korea, Cambodia, Vietnam, and even Japan.

In Central Asia, the Tang destroyed the Eastern Turkic Khaganate and subjugated the Western Turkic Khaganate in the mid-seventh century, breaking the power of the Göktürks and establishing a protectorate in the Tarim Basin. It was inevitable that they would bump up against the expanding Arabs, who entered the region in the early 8th century (they captured the city of Samarkand, for example, in 710). The two empires clashed for the first time in 715. The Umayyad caliphate deposed a local ruler in the Fergana Valley (modern Uzbekistan with parts of Tajikistan and Kyrgyzstan) and installed a client in his place. The ruler they’d ousted appealed to the Tang for assistance, and the Tang eventually restored him to the throne. Over the next couple of years, the Umayyads made diplomatic inroads with the Tibetan Empire as well as with the Türgesh, a Turkic confederation that emerged from the collapse of the Western Turkic Khaganate. Along with their new allies, the Umayyads besieged a couple of cities in the Tarim Basin (modern Xinjiang), but were defeated by the Tang so severely that the Türgesh decided it would be sensible to switch sides and help chase the Arabs out of the region—though the Arabs eventually got the better of them.

In exerting control over Central Asia, and especially in fending off the Umayyads, the Tang relied to a great extent on mercenaries hired from the Karluk peoples. The Karluk were another Turkic confederation that had lived under the Göktürk khanates in the region of the Altai Mountains (which span parts of modern China, Kazakhstan, Mongolia, and Russia). Chinese leaders, playing the long game, cultivated a relationship with the Karluks that came to fruition when the latter finally agreed to turn on the Göktürks, after which point they served as Chinese vassals and proxies. Their involvement at Talas will be crucial to the final outcome.

As I mentioned earlier, the Abbasid Revolution overthrew the Umayyads in 750, and the new caliph, al-Saffah (d. 754), decided to send his forces out to secure the new dynasty’s control over the caliphate. Another confrontation with the Tang was, again, inevitable. It happened at…well, that’s complicated, but at least we do know that the battle took place on…well, no we don’t really know that either. The Battle of Talas took place on the frontier of two great empires, which means it wasn’t at the center of any empire and therefore the level of detail available in extant sources is not great.

It seems nearly certain that the battle took place. While not well attested in Arab sources there are mentions of it in some chronicles, and it’s better attested the official court histories of the Tang Dynasty. Chinese chroniclers probably had no reason to invent a fictional battle that (spoiler alert) their side lost, although we can’t be completely sure about that because inter-dynastic politics may have crept into the historical record. We can be a bit less certain (but still pretty certain) that the battle took place in 751. We can say with some confidence that it took place in the summer of that year in the region of the Talas River, which runs from the Talas region of Kyrgyzstan west into Kazakhstan. Beyond that, it’s pretty murky. So I’m posting this today because it’s mid-summer, which fits the available evidence, and because I feel like it.

In terms of numbers, we’re at the mercy of unreliable Chinese sources, who put the Arab army in the ridiculous neighborhood of 200,000, and Arab sources, who put the Chinese army in the nearly as ridiculous neighborhood of 100,000. A tentative consensus has emerged that the Chinese-Karluk army numbered around 30,000, with two-thirds of it made up of Karluks (a roughly 2-to-1 ratio of Karluks to Chinese soldiers seems to be widely accepted regardless of what the total number of soldiers was). The Abbasids, again joined by the Tibetans, probably outnumbered them and may have outnumbered them by a large margin, but any specific figures you might see are mostly guesses.

Assuming the battle really did take place, we know that its immediate spark involved the small kingdom of Shash (around the modern city of Tashkent), whose ruler was killed by Tang forces and whose son and heir appealed to the Abbasids for help. In this sense, Talas mirrored the first contact between the Tang and the Umayyads, after the Umayyads ousted the ruler of Fergana and he appealed to the Tang for help. We know the battle lasted a few days, though some sources claim it lasted four days while others claim it was five. And we know that the outcome in this case was an overwhelming Abbasid victory. The key factor in the victory was apparently a mid-battle face (or heel, depending on your point of view) turn by the Karluks, who began attacking the smaller Tang contingent in their army. However large that contingent was, it’s believed only a small fraction of its soldiers survived to make the long trip home.

Unlike the previous two encounters between the Arabs and the Tang in Central Asia, it’s tempting to say that the Battle of Talas was decisive, since it was after this engagement that Islam really began to take off among the Turkic peoples of the region. But this is more correlation than causation. Chinese influence in Central Asia did decline precipitously in the 750s, but that’s because of the An Lushan Rebellion, which ran from 755 to 763 and ended with the Tang still in power, but greatly weakened and unable to hold on to much of their imperial periphery. Certainly the Arab victory at Talas helped secure them a place in Central Asia, but it was the later rebellion that enabled them to push deeper into the region.

The Tang retrenchment caused by the rebellion is what really opened the gates for Islam to spread among the Turks, and of course the results are pretty apparent today. The introduction of Islam to Central Asia on a large scale also may have contributed to the evolution of a Chinese/eastern Buddhism independently of Indian Buddhism, as it left eastern Buddhists somewhat isolated from their South Asian coreligionists. I don’t know enough about the development of Buddhism to really comment on this, but it is true that innovations like Zen, a syncretization of Buddhism and Taoism, start appearing in the later Tang period, after An Lushan.

The rebellion also, as an aside, led to the introduction of Islam into China itself. The Tang Emperor Suzong (d. 762) appealed to the Abbasids for help putting down the rebellion and they obliged by sending a few thousand fighters east (whatever diplomatic rupture Talas may have caused was clearly patched up pretty quickly). Many of those fighters later settled in China and created the first Islamic communities in what had hitherto been a land so distant and unknown to the Arabs that writers would often use the word “China” less as a real place name than as a term meaning a far off, exotic land.

According to one very traditional narrative, the Battle of Talas had one rather far-reaching effect that had nothing to do with the geopolitics of Central Asia. We’re told that among the Chinese prisoners taken by the Abbasids during the battle were men who were knowledgeable in paper-making. They brought this knowledge back with them to Samarkand, and from there it was transmitted to the Middle East and on to Europe. This story sounds cool, and it really stuck with me when I learned about it in grad school. There’s even some general support for the idea that Chinese prisoners of war and other travelers to the Islamic world brought their crafts with them, including a travelogue (only a few excerpts from which survive) by one of the Chinese soldiers captured at Talas, a man named Du Huan.

The thing is, and I take no pleasure in saying this, that the evidence for the specific transmission of paper-making to the Islamic world after Talas is…paper thin (I’m very sorry). The main glitch is that there’s clear evidence of paper-making in Central Asia well before Talas, as Mark Kurlansky points out in his 2016 book helpfully titled Paper:

The problem with this story is that paper had already been in use in Central Asia in the fourth century, as was proven in 1907 by the legendary Hungarian-born British archaeologist Marc Aurel Stein. Stein discovered a packet of undelivered letters written on paper in Central Asia dating from 313-314. An angry wife whose husband had gone off somewhere and left her with no money had written them. “I would rather be a dog’s wife or a pig’s wife than yours,” she wrote.1

Aside from giving us a little glimpse at fourth century marital bliss, the letters show pretty conclusively that Chinese prisoners did not introduce paper to Central Asia in 741. Chinese accounts further have paper being manufactured in the region by at latest the fifth century. Furthermore, Central Asian paper-making techniques evolved differently from Chinese techniques and it was the Central Asian style—specifically the kind of paper produced at Samarkand—that made its way west into the Arab world. That transmission of technology was a seminal event in the development of Islamicate civilization, but it seems there’s little reason to believe it had anything to do with the Battle of Talas.


Kurlansky, Mark, Paper: Paging Through History (W.W. Norton and Company, 2016), p. 49